Lung cancer is the most common cancer in the world. It is a highly lethal disease in women and men, and new treatments are urgently needed. Previous studies implicated a role of estrogens and estrogen receptors in lung cancer progression, and this steroidal growth-stimulatory pathway may be promoted by tumor expression and activity of aromatase, an estrogen synthase. We found expression of aromatase transcripts and protein in human non–small cell lung cancer (NSCLC) cells using reverse transcription-PCR and Western immunoblots, respectively. Aromatase staining by immunohistochemistry was detected in 86% of archival NSCLC tumor specimens from the clinic. Further, biological activity of aromatase was determined in NSCLC tumors using radiolabeled substrate assays as well as measure of estradiol product using ELISA. Significant activity of aromatase occurred in human NSCLC tumors, with enhanced levels in tumor cells compared with that in nearby normal cells. Lung tumor aromatase activity was inhibited by anastrozole, an aromatase inhibitor, and treatment of tumor cells in vitro with anastrozole led to significant suppression of tumor cell growth. Similarly, among ovariectomized nude mice with A549 lung tumor xenografts, administration of anastrozole by p.o. gavage for 21 days elicited pronounced inhibition of tumor growth in vivo. These findings show that aromatase is present and biologically active in human NSCLCs and that tumor growth can be down-regulated by specific inhibition of aromatase. This work may lead to development of new treatment options for patients afflicted with NSCLC. (Cancer Res 2005; 65(24): 11287-91)
Lung cancer is the leading cause of death from cancer in both women and men in the United States. In view of the recent surge in female death rates from lung cancer (1), the potential role of both exogenous and endogenous estrogens in lung cancer development, especially adenocarcinoma in women, requires further investigation (1–3). Estrogens contribute to differentiation and maturation in normal lung (4) and also stimulate growth and progression of lung tumors (3, 5). These biological effects are mediated by estrogen receptors (ER), with ER transcripts and proteins reported in most non–small cell lung cancers (NSCLC; refs. 3, 5–10). Moreover, as in breast, aromatase, a cytochrome P450 enzyme complex, mediates synthesis of estrogens in lung tissues (5, 11), and local production of estrogens in women and in men could affect lung tumor progression in ER-expressing malignancies (5, 12). Potent third-generation aromatase inhibitors, such as anastrozole, interact competitively with heme groups of cytochrome P450 components of aromatase. As in breast cancer (12), aromatase inhibitors may prove useful to block estrogen synthesis and estrogen-induced growth in lung cancers. Thus, we investigated the role of aromatase in lung cancer by assessing expression and biological activity of the estrogen synthase in lung cancers and we confirmed tumor expression of ER. In addition, antitumor efficacy of anastrozole was tested in experiments with human lung tumors growing in vitro and as xenografts in nude mice.
Materials and Methods
Cell lines and reagents. NSCLC cells H23, A549, A427, A125, H2122, H1299, H157, H647, H460, H1650, H3255, and H226, and control breast cancer cells SKBR3, MCF-7, and T47D were from the American Type Culture Collection (Manassas, VA). Cells were maintained as before (5, 13). Anastrozole was from AstraZeneca (Wilmington, DE) and TeamLife Research (New York, NY), whereas testosterone and androstenedione were from Steraloids (Newport, Rhode Island) and Casodex (bicalutamide) was from AstraZeneca.
Antibodies and Western blotting. Antiaromatase antibody (CYP19) C-16 was from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies to ERα (1D5) and ERβ (D7N) were from Invitrogen/Zymed (Carlsbad, CA). Cells were prepared for electrophoresis and Western blotting as before (5, 13).
RNA isolation and reverse transcription-PCR assay of aromatase. Total RNA was extracted from cells using TRIzol reagent (Invitrogen) as before (14). In brief, 2 μg total RNA were reverse transcribed to first-strand cDNA with SuperScript II reverse transcriptase (Invitrogen). PCR was done with Platinum Taq polymerase using the “Hot Start” method (95°C/3 minutes) followed by 40 cycles of 94°C/15 seconds, annealing of 60°C/30 seconds (for aromatase), and 50°C/30 seconds [for glyceraldehyde-3-phosphate dehydrogenase (GAPDH)], 68°C/30 seconds, followed by a final extension of 68°C for 10 minutes.
Aromatase activity in cell lines and archival tumor specimens. Aromatase activity was assessed by use of radiolabeled substrate, [1β-3H]androst-4-ene-3,17-dione (Perkin-Elmer, Boston, MA) with established methods in lung cells (16) as well as in previously frozen lung NSCLC specimens (17). Controls included use of cells or tissues without [1β-3H]androst-4-ene-3,17-dione and deletion of cells or tissues.
As aromatase catalyzes conversion of androstenedione and testosterone to estrone and estradiol, respectively, its activity was determined by production of estradiol after treatment with testosterone. Cells were plated in RPMI medium with 10% fetal bovine serum (FBS). After reaching 50% confluence, cells were rinsed with PBS and incubated with phenol red-free medium with 1% dextran-coated, charcoal-treated FBS for 48 hours to deplete estrogen (5, 13). Thereafter, medium was refreshed and cells were treated with testosterone for 48 hours. Estradiol production was measured with ELISA (ALPCO Diagnostics, Windham, NH) using the instructions of the manufacturer in four independent experiments.
Immunohistochemical assays. Formalin-fixed, paraffin-embedded lung tumor samples from 53 patients were obtained, and immunohistochemistry was done by established methods (5). Negative controls (deletion of primary antisera as well as IgG isotype control) and positive controls (tissues with known aromatase) were prepared in parallel.
Cell proliferation assays. Lung tumor cells were plated in RPMI medium with 10% FBS. After cells reached 50% confluence, they were rinsed with PBS and incubated with phenol red–free medium with 1% dextran-coated, charcoal-treated FBS for 48 hours (13). Then, the medium was refreshed and cells were treated with anastrozole. In addition, androstenedione or estrogen was added to some cells without drugs. After 72 hours, cells were counted manually (13).
Inhibition of aromatase in cell lines. A549 cells were plated and grown to 50% confluence (13); cells were then washed and treated with anastrozole for 48 hours. Aromatase activity was assessed in triplicate (16), with untreated A549 cells used as controls.
In vivo tumor xenograft studies. Ovariectomized nude mice at 6 weeks of age were from Harlan Sprague-Dawley (Indianapolis, IN). Animals were housed in a pathogen-free environment with controlled light and humidity and received food and water ad libitum. To prepare lung tumor xenografts, A549 cells (2 × 108 per mouse) were injected s.c. with 1 tumor/mouse (13). Androstenedione (0.1 mg/mouse) was injected s.c. daily in each mouse throughout the experiment to provide substrate for aromatase. After tumors reached 50 to 75 mm3, mice were randomized into uniform groups based on animal weight and tumor volume for treatment with control or anastrozole (13, 18). Anastrozole was administered by daily p.o. gavage at 0.1 mg/kg (20 μg/mouse) in water. Treatment duration was 21 days. Tumor volumes were measured every 3 to 4 days, with tumor volume calculated by (l × w × h), where l is tumor length, w is tumor width, and h is tumor height in mm.
Statistical analyses. Data were analyzed using ANOVA and Student's t tests (18). Tests of statistical significance were two-sided, with differences significant if P < 0.05.
Clinical samples. Patient specimens and information were collected under Institutional Review Board–approved and Health Insurance Portability and Accountability Act–compliant protocols at University of California at Los Angeles Medical Center.
Aromatase and estrogen receptor expression in non–small cell lung cancer cell lines. Expression of ERα and ERβ proteins in lung tumor cells was assessed by Western immunoblotting, with results for six representative lung cell lines and for selected breast cancer controls shown in Fig. 1A. The classic 67 kDa ERα, as well as a 46 kDa variant, occurs in MCF-7 breast cells. The several lung cancer cells also express these ERα forms, albeit at lower levels than breast cells. Lung cells also contain a larger 80 kDa ERα form that may represent another variant with an in-frame duplication of exons 6 and 7 (2). As shown in Fig. 1B, ERβ forms at predominantly 59 kDa were found in lung cancer cells, as well as in breast cells (2, 5).
Expression of aromatase protein in lung tumor cells was assessed, with results for six representative cell lines shown (Fig. 1C). As in SKBR3 breast cancer cells with known high aromatase expression, the enzyme occurred in most lung cancer cells, with molecular size ∼54 kDa (12). To confirm these findings, reverse transcription-PCR (RT-PCR) was used to detect aromatase transcripts in lung cells. As shown in Fig. 1D, a 294 bp transcript was detected in all lung cancer cells examined, consistent with earlier reports (15). A similar RT-PCR product was detected in SKBR3 cells with high aromatase levels.
Aromatase activity in human lung tumors. To assay aromatase activity, we measured conversion of androstenedione substrate to estrogens (16). Activity occurred in all lung cancer cell lines tested and ranged from 0.37 to 3.21 pmol/mg/h, with average aromatase activity of 1.49 ± 0.23 pmol/mg/h (see Fig. 2A). As expected, SKBR3 cells had high activity, 5.84 pmol/mg/h, a level exceeding that of most breast cells. In our study, MCF-7 breast cells had <5% of SKBR3 cell activity (not shown), as in previous reports (16). Among a total of 12 human NSCLC cell lines evaluated, aromatase activity was highest in H2122, A549, and H647 (not shown) cells and lowest in H23 cells (Fig. 2A).
Aromatase activity was also quantitated by estradiol production after treatment of A549 cells with testosterone at 1 to 10 nmol/L (Fig. 2B). Increased testosterone levels promoted a significant increase in estradiol biosynthesis in NSCLC cells, providing further evidence for functional aromatase (P < 0.001).
To assess aromatase activity in clinical material, archival lung tumor specimens were tested by radioassays with [1β-3H]androst-4-ene-3,17-dione substrate (17). Paired tumor and neighboring normal tissue was examined from 10 donors (Fig. 2C). Aromatase activity in normal tissue averaged 0.081 ± 0.052 pmol/mg/h, whereas corresponding tumors had significantly higher activity, averaging 0.15 ± 0.03 pmol/mg/h (P = 0.001; Fig. 2C).
Immunohistochemical assay for aromatase in lung tumors. To determine the prevalence of aromatase in human NSCLC, we assessed 53 archival tumor specimens by immunohistochemistry. Representative examples of specific staining patterns of different tumor types classified as aromatase-positive, including a positive control breast cancer specimen, are shown in Fig. 3A. Staining was primarily in tumor epithelial cells, with distinct clusters of positive-staining tumor cells throughout (Fig. 3A). Often, adjacent normal epithelium of bronchioles also exhibited weak aromatase staining, but supporting stromal tissues had little aromatase. Macrophages contained aromatase in some samples, about twice as often in females as in males. Aromatase staining was not observed in lymphocytes or interstitial surrounding tissue. Using a cutoff score of 15% total cell staining for aromatase positivity, tumor specimens were classified as negative (≤15% stained) or positive (>15% stained), with examples of aromatase-positive lung tumors in Fig. 3A. Using this evaluation system, 86% of lung tumor specimens expressed aromatase (see summary; Fig. 3B). This finding was relatively uniform among different lung NSCLC histologies evaluated, with 88% of adenocarcinoma and 87% of squamous carcinomas positive for aromatase. Of the limited number of bronchioalveolar and adenosquamous samples available, 75% and 100%, respectively, scored positive (Fig. 3B). Using these scoring criteria, 88% of female and 86% of male NSCLC samples were positive.
Aromatase inhibitor blocks lung tumor growth in vitro and decreases aromatase activity. As reported previously, both A549 and H23 lung tumor cells express ER (2, 5, 9). To assess estrogen dependence of lung cells for growth, cell proliferation in response to estrogen and androstenedione, an aromatase substrate to promote cellular estrogen production, was tested. As shown in Fig. 4A, increasing doses of estrogen from 1 to 20 nmol/L elicit enhanced proliferation of A549 cells. At 20 nmol/L, estrogen induced a significant 2.4-fold increase in cell proliferation (P < 0.001) as before (2, 5). Similarly, androstenedione at increasing doses significantly stimulated proliferation of A549 cells (Fig. 4A). At 100 nmol/L, androstenedione elicited a 2.1-fold increase in A549 cell growth (P < 0.001).
To assess direct effects of an aromatase inhibitor on lung cancer cell proliferation, tumor cells were treated with or without anastrozole in vitro. Anastrozole significantly reduced cell growth in a dose-dependent manner (P < 0.001; Fig. 4B). Fifty percent growth inhibition occurred at 0.1 mmol/L anastrozole for A549 and at 3.5 mmol/L anastrazole for H23 cells, suggesting that tumors with both low and relatively higher levels of aromatase are sensitive to aromatase inhibition (Fig. 4B).
The ability of anastrozole to suppress aromatase activity in lung cells was also tested (Fig. 4C). A549 cells were treated 48 hours with anastrozole and aromatase activity was then determined and compared with controls. A significant dose-dependent decrease in aromatase activity was found, with more pronounced aromatase inhibition at 10 to 50 μmol/L anastrozole (P ≤ 0.01). This dose range corresponds well with concentrations noted to be more effective in growth inhibition in vitro (Fig. 4B).
Aromatase inhibitor suppresses growth of lung tumor xenografts in vivo. To assess in vivo antitumor effects of anastrozole, A549 cells that express aromatase and ERβ (2, 5, 9) and respond well to estrogen (2) were grown as xenografts in ovariectomized nude mice treated daily with aromatase substrate, androstenedione. After tumors grew to a limiting size of 50 to 75 mm3, mice were divided into two treatment groups, control vehicle and anastrozole, and then treated for 21 days. Results in Fig. 4D show that, by day 21, there was exponential growth of tumors in control mice, whereas growth of tumors in mice treated with anastrozole was significantly suppressed by >90% compared with controls (P < 0.001).
The current data and previous findings suggest that treatments targeted to estrogen signaling pathways in lung cancer may have previously unsuspected antitumor efficacy (2, 5, 9, 10, 19). In nonreproductive tissues, the lung possesses among the highest levels of ER transcripts, especially those for ERβ (7). Our results confirm that wild-type and variant forms of ERα and ERβ occur in lung tumors (2, 5, 8, 9, 19) and participate in growth promotion in lung cancer (2, 5, 19). Moreover, we show with different methods that local production of estrogens occurs in lung tumor cells and NSCLC tissue samples through the action of aromatase, and ligands generated may then bind and activate ER. Aromatase activity occurs in most tumors from male and female patients with different NSCLC histologic types. The majority of lung tumors in this study were adenocarcinomas, the most common histologic type of lung cancer in young persons, women of all ages, and never smokers of both sexes (1). Further, levels of aromatase activity were significantly greater in tumors compared with those in nearby normal tissue. Similarly, among archival lung NSCLC evaluated by immunohistochemistry, aromatase localized predominantly in tumor epithelial cells, with less staining in surrounding stromal tissues. Of note, aromatase also occurred in normal epithelial cells distant from primary tumor and in some infiltrating macrophages. In breast, macrophages constitute another source of estradiol to promote cell proliferation (20).
To further assess the notion that estrogen signaling promotes lung cancer progression, antitumor activity of aromatase inhibitors was assessed. Anastrozole, when administered daily to postmenopausal women with breast cancer, inhibits in vivo aromatization by 96.7% and clinical trials show strong antitumor efficacy of anastrozole in breast cancer (12, 16). Similarly, in our study, lung aromatase activity was significantly reduced by anastrozole, and treatment of lung tumor cells in vitro and xenografts in vivo with anastrozole elicited a corresponding suppression of tumor growth compared with controls. Therapeutic targeting of lung cancer to block specific receptor pathways, such as estrogen signaling, may provide new options for patients afflicted with lung malignancies.
Grant support: NIH Lung Cancer Specialized Programs of Research Excellence Program grant P50-CA90388 (S. Dubinett, M. Fishbein, L. Goodglick, and R. Pietras), Stiles Program in Oncology (in vitro data; R. Pietras and D. Marquez), Howard Hughes Medical Institute Fellowship (O. Weinberg), Piansky and Reiss Trusts (M. Fishbein) and EDRN CA-86366 (L. Goodglick).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Samuel Olson for technical assistance.